The Economic and Social Review, Vol. 17, Mo. 4. July, 1986, pp. 293-314

Denial, Polarisation and Massacre: A Comparative Analysis of Northern Ireland and Zanzibar

J. L. P. THOMPSON Center for the Social Sciences, Columbia University

Abstract: The Zanzibar Revolution ofl964 and the Northern Ireland conflict since 1968 are compared. They are similar in their polarisation processes, but differ in the level of killing, which is much higher in Zanzibar. Northern Ireland has, however, experienced the tinge of massacre. Denial of the severity of the ethnic conflict is documented in both cases, and its impact on polarisation and the level of killing explored. It promotes polarisation by precluding the application and development of the ability to negotiate and regulate conflict; and it facilitates massacre by preventing its control by the public or the security forces.

I INTRODUCTION olarisation is "a process of increasing aggregation of the members of the P society into exclusive and mutually hostile groups, entailing elimination of the middle ground and of mediating relationships" (Kuper, 1977, p. 128). Massacre occurs to the extent that individuals are killed on the basis of member­ ship in a collectivity. It may be conceptualised as a variable which ranges from no victims at the low pole through the killing of isolated individuals to, at the upper extreme, the massacre of large numbers of people who are members of a particular social and/or political category.1 Polarisation typically precedes massacre, but does not necessarily result in it. In this paper I seek explanations of two paradoxical and disturbing phenomena. The first is that after periods of

1. This conception is adequate for purposes of the present paper. I have proposed more detailed measures of severity elsewhere (Thompson, 1985a, 1985b).

*I am indebted to Christopher Hewitt, Tom Garvin, Tormod Lunde, Robert V. Robinson, M. G. Smith, Murray Strauss, Dennis Wrong, and an anonymous referee for comments on previous drafts of this paper, and to the members of a seminar on political violence organised by the Harry Frank Guggenheim Foundation, New York, for their valuable input. Financial assistance from the Leverhulme Trust is gratefully acknowledged. relative quiescence, societies can suddenly embark on periods of rapid polarisa­ tion and conflict. The resurgence of the Northern Ireland conflict in the late 1960s is a case in point. The second paradox is that "normal" people — who are not social outcasts, not psychologically abnormal, and have no previous personal history of violence — may stand by or participate in massacres com­ mitted in their own society. In Nazi Germany, millions must have had some awareness that the genocide against the Jews was occurring, but few protested.2 And in the Lebanese conflict:

Priests tortured, as did devout Muslims. Young girls of the best Christian society, petty bourgeois costumed at Pierre Cardin or Courreges, admirers of Brassens and Bob Dylan, castrated prisoners; university faculty, advocates of coexistence between the communities, embodying the wisdom of Islam and of Christianity, gouged out eyes and disembowelled women (Desjardins, 1976, p. 39, quoted in Kuper, 1981, p. 104).

How can these things be? The question is so troubling that it is often avoided. Here it is explored through a comparative analysis of the Zanzibar Revolution of 1964 and the Northern Ireland conflict since 1968. The first half of the paper provides a summary introduction to Zanzibar, establishes its comparability with Northern Ireland, and compares the polarisation process and level of violence in the two societies. The latter half examines the impact of denial on polarisation and massacre.

II ZANZIBAR

Zanzibar consists of two islands in the Indian Ocean which lie near to the coast of East Africa. Cyclical winds have historically exposed it to immigrants from Arabia and India, and given it strategic importance. In 1830, after some centuries of intermittent domination by Arabs, and a Portuguese interlude, it became an independent Arab state headed by a Sultan. This regime exercised sovereignty over towns and trading centres on the African mainland, and was a centre for the slave trade. After a brief period as the most powerful political force in the Indian Ocean it came increasingly under the influence of European, and particularly British, expansionism. Under a Protectorate agreement signed by

2. "... as the process unfolded, its requirements became more complex and its fulfilment involved an ever larger number of agencies, party offices, business enterprises, and military commands." ... "the very nature of administrative planning, of the jurisdictional structure, and of the budgetary system precluded the special selection and special training of personnel .... However we may wish to draw the line of participation, the machinery of destruction was a remarkable cross-section of the German population. Every profession, every social skill, and every social status was represented in it" (Hilberg, 1961, p. 640, p. 649). the governments of Great Britain and Zanzibar in 1890 the British in theory shared sovereignty with the Sultan, but in practice dominated. However, until their departure in 1963 British officials conceived of the society as an Arab nation which would evolve into a constitutional monarchy with the Sultan as Head of State. They accorded preferential treatment to Arabs for legislative and administrative positions, and prevented other communities from challenging the Arabs' pre-eminent position (Lofchie, 1965, Chapter 2). In 1958 the population of the larger of the two constituent islands, itself named Zanzibar, was 166,000, and of the smaller, Pemba, was 134,000.3 The ethnic composition of the society is summarised in Table 1. Some one-sixth of the population were Arabs. Shirazis, blacks indigenous to Zanzibar who did not identify themselves as Africans, constituted over half the population. They were predominantly engaged in agriculture and fishing. The concept of Shirazi identity denotes Afro-Persian admixture, originally caused by the thirteenth century influx of Persians from the principality of Shiraz who were absorbed by the indigenous communities (Lofchie, 1965, pp. 24-25; 1969, p. 285). In the mid-twentieth century it differentiated blacks indigenous to Zanzibar from recent African immigrants. The latter — "Mainland Africans" — constituted one-fifth of the population. Mostly manual workers, they were concentrated in Zanzibar town, on the larger island, and not well integrated into the society. The small Asian community dominated wholesale and retail trade and the middle tiers of the civil service.

Such a summary neglects both integrating elements and diversity within the major groups. While there was some politically significant religious differentia-

Table 1: Racial and ethnic groups in Zanzibar, 1948 Group Number Percentage

Arabs 44,560 16.9 Shirazis 148,480 56.2 Mainland Africans 51,380 19.5 Asians 15,211 5.8 Others 4,531 1.7

Total 264,162 100.1

Source: Zanzibar Protectorate 1953, Tables I, XV.

3. Until 1963 the domain of the Sultan also included a ten-mile strip of the African mainland, which was rented by the British and administered as part of Kenya. tion, Zanzibar was largely homogeneous in terms of religion, over 95 percent of its inhabitants being Muslims, and religion was a substantial integrating force.4 None of the ethnic sections was internally homogeneous. Mainland Africans originated from a wide variety of East African tribes (Zanzibar Protectorate 1953, p. 108; Lofchie, 1965, p. 82); some Asians had become wealthy land­ owners, and Shirazis were differentiated into Hadimu, Timbatu and Pemba, although by the mid-twentieth century these distinctions had lost a good deal of their force. The Arab section ranged from established elite families through a peasant class of small farmers and rural shopkeepers (the bulk of the Arab population) to Manga Arabs, low-status recent immigrants from Oman whose relationship to the Arab elite was openly hostile (Lofchie, 1965, pp. 77-79). Nevertheless, Arab identity was prestigious, and members of other sections could and did pass as Arabs. These included, but were not limited to, the low- status Swahili, who were stigmatised as being of slave descent (Lofchie, 1965, pp. 73-77; Bennett, 1978, p. 268). This porosity of ethnic boundaries was a some­ what general feature of Zanzibar society. Africans of mainland origin who became acculturated were regarded as Shirazi (Kuper, 1977, p. 149), and after the revolution of 1964 many who outwardly appeared to be of African origin and had passed as Arabs reverted to an African identity (Martin, 1978, p. 25). It has been suggested (Bennett, 1979, p. 268) that the British might have ended ethnic difficulties in the Protectorate by encouraging the tendency for lower-status groups to pass as Arabs. This is perhaps possible, given the example of Mexico, where extensive passing did result in substantial homogenisation of the popula­ tion (van den Berghe, 1978, pp. 42-58). Contrary to a view which we will encounter below, a tradition of violence was established in Zanzibar society long before the Revolution. This is stated explicitly by Lofchie (1965, p. 205). whose authoritative account refers to many manifestations of it. Thus in 1828 Sultan Seyyid Said had encouraged Hadimu acceptance of the legitimacy of his regime by positioning his entire fleet, including numerous warships and transport carriers for several thousand soldiers, directly offshore. A slave rebellion in 1840 proved beyond the control of

4. According to Lofchie (1965, pp. 12-13):

Islam in Zanzibar created a pervasive religious environment highly favourable to inter-racial political solidarities. This environment furnished Arab leaders with the basic ingredient of an eifective appeal for African political support: a common faith. Arab and African Zanzibaris shared not only the same theology, but all the various institutions and practices which accompanied it such as mosques, Koranic schools and a host of identical holidays, rituals and ceremonies. These made Islam a highly visible symbol of the common religious identity of Zanzibaris of all races.

Lofchie adds that the Koran is explicit in making harmonious race relations a religious duty, and that the Arab leaders in Zanzibar employed these precepts to foster a widespread conviction among Africans that the Faith enjoined multiracial unity upon all believers as a holy obligation. the substantial local Arab forces, and continued for six months until military reinforcements could be brought in from Arabia. The expropriation of Hadimu land in the 19th century involved violence. The Sultan's caravans into the East African interior were suffciently heavily armed to be described as constituting the first real invasion of the area. In 1893 the British secured the succession of the least intractable of several contenders for the Sultanate by force. A "sort of private blood feud" between Omani and Manga Arabs erupted into violence and rioting in 1928. In 1936 there were riots over the imposition of quality controls on cloves and copra. In 1938, when the political atmosphere "bordered on hysteria and racial violence", Asians instituted a clove boycott against Arabs, and armed bands of the latter roamed the rural areas intimidating Asian shop­ keepers, hoping to pressure the Asian community to abandon the boycott. A land crisis sparked riots in 1951.' In the 1950s this tradition was imported into a new arena, as political conflict developed between emergent nationalisms over the form which Zanzibar society was to take after the departure of the British. This clash was reinforced by pre­ existing, and persisting, economic antagonism between the ethnic groups. The main protagonists in the escalating conflict were the Arab and Mainland African sections. Both demanded national independence, but the former feared African domination, and the latter sought an African majority rule which would preclude domination by Arabs. Africans were hindered by poor organisation, economic disadvantage, and low educational resources, and a British decree banning political involvement by civil servants, who constituted one of the few pools of potential black leaders. The Arabs were first in the field with a political party, the ZNP (Zanzibar Nationalist Party), which elaborated a specifically Muslim Zanzibari nationalism that embraced the Shirazi but excluded Main­ land Africans. An African nationalist party, the ASP (Afro-Shirazi Party), which mobilised in reaction analogously included Shirazis but excluded Arabs. Each of these parties sought a coalition which would take the political spoils in the new democratic system while excluding its key ethnic antagonist. The Shirazis, while courted by both, were often reluctant to respond, particularly on Pemba island.6 The process of polarisation intitially centred on Zanzibar island, although it later spread to Pemba. As the society became intensely politicised and the antagonism of the two main contending parties more intense, outbreaks of

5. For these events see Lol'chie (1965, pps. 35, 41, 46, 50, 57, 121, 149-150, 205).

6. According lo Lofchie (1965, p. 172) "The disproportionate political influence of these immigrant minorities [i.e. Arabs and Mainlanders] engendered among the Pemba Shirazis a sense that they were innocent bystanders in a species of cold war waged on their soil by alien elements." violence grew more severe.7 In 1955, in a highly charged atmosphere, the Sultan was assassinated by a demented Arab. Landlord-squatter conflict led to violence, evictions, and fears of reprisals between Arabs and Africans in 1957. In 1958 "persistent outbreaks of disorder threatened to engulf the society in chronic racial warfare" (Lofchie, 1965, p. 188). By 1959 youthful supporters of both major parties participated in armed paramilitary youth wings which baited and assaulted members and leaders of the opposite party. The elections of 1961 pre­ cipitated several days of uncontrollable rioting on Zanzibar Island. Bands of Africans roamed the plantation area, looting, pillaging, and murdering. Political candidates were beaten, and 68 people died in riots, 64 of them being Arabs. The actual revolution, which occurred in January 1964, just after the withdrawal of the British, was a decidedly "freelance" affair, although it unleashed violence far beyond what was necessary to effect a seizure of power. Its leader, John Okello, was subsequently removed to the mainland by remaining politicians. Other coups and plots were in the wind; Okello happened to be the one who initiated the coup which allowed those inclined to massacre Arabs to do so. Some 5,000 of the latter were killed,8 and there was a total decrease in the Arab population, through murder, repatriation, and emigration, of some 10,000. Zanzibar was reconstituted with a specifically African ethos. In 1964 it united with Tanganyika, its larger neighbour, to form the state which was soon named Tanzania. The Zanzibar revolution was the culmination of an historical process in a society in which a certain, although not particularly high, level of economic and ethnic violence was endemic. While the British contributed to this by buttressing and protecting the Arab section, the Arab/African conflict predated British involvement, and before the revolution the British were a restraining force (Kuper, 1977, p. 156-157). The revolution is often referred to as simply a revolt of Africans against Arabs, which it was. But it also involved massacre of parts of the Arab population.9 While the Arabs are sometimes described as an "alien oligarchy", this is a delegitimating stereotype rather than an accurate descrip-

7. For the political developments see Lofchie (1965, pp. 127-281). Kuper (1977, pp. 145-170) focuses on the polarisation process and violence.

8. Bennett (1979, p. 267) regards this as a conservative estimate. Lofchie gives an estimate of 3,000 fatalities.

9. Even the patently self-serving account of Okello acknowledges massacre, as in:

... in many cases, I ordered my men to burn Arab houses rather than kill the people. But the bitterness of years of oppression inevitably produced many acts of vengeance; one day I accompanied my men to round up 41 Arab men who were hiding with 18 women and 10 children. I had left them under guard to be collected by a lorry and taken to Zanzibar town for detention; but when the lorry arrived they had all been killed (Okello, 1967, p. 153).

For a survivor's account referring to other such incidents see Kharusi (1967). tion. They were not aliens in any meaningful sense,10 and the vast bulk of Arabs were not oligarchs. It is intriguing, given their use of violence in past history, that the Arabs were much more the victims than the instigators of violence in the revolution.

Ill COMPARING THE CASES"

Both Northern Ireland and Zanzibar are plural societies, characterised by severe racial, ethnic, or religious conflict.12 They fall, moreover, within that subset which displays two distinct tiers of domination. Some of this subgroup, such as Zanzibar, Rwanda, and Burundi, were already plural before a new power imposed its overall domination: in others, such as Northern Ireland, the metropolitan power introduced a new section which either was or became ethnically distinctive. But whatever their genesis, plural societies with two tiers of domination are typically highly complex in social structure, and often charged with a high potential for violence and even genocide (Kuper, 1981, pp. 61-73). Our two cases are similar in several specific respects:

1. Their de facto relationship to the United Kingdom. In constitutional terms, Northern Ireland's position as part of the United Kingdom contrasts with Zanzibar's as a sovereign independent state capable of concluding treaties, such as the Protectorate Treaty, with other states. But substantively the United Kingdom, a larger power unable in the end (in Zanzibar) and so far (in Northern Ireland) to impose an enduring pax Britannica, dominates in each case. (Interestingly, when Zanzibar amalgamated with Tanganyika to form Tanzania, their constitutional relationship in the new entity was modelled on that between Great Britain and Northern Ireland (Lofchie, 1965, p. 280)).

10. Bennett (1979, p. 252) cautions against regarding the Arab section as "alien" to Zanzibar society:

Zanzibar's Arabs, many of them the issue of families resident in the islands for generations, were as much integral members of the local society as Africans whose families were of similar local ancestry. The Arabs of Zanzibar and Pemba were no more alien in their country than are the black American inhabitants of the United States: both groups differed in some ways from the more numerous elements of the population of their native lands, but both were citizens sharing equally most of the common characteristics of their nation's makeup.

11. For relevant discussions of comparative analysis see Lijphart (1971) and Frcdrickson (1981, pp. xiii-xx). Since accounts and summaries of the social structure, historical background, and tradition of political violence in Northern Ireland are readily available (e.g. Darby, 1976 , 1983; Buckland, 1981; Farrell, 1980; Stewart, 1977), these topics were not reviewed here.

12. For the concept of the plural society see Kuper and Smith, (1969) and for its application to Northern Ireland and Zanzibar see, respectively, Thompson, (1983) and Lofchie, (1969). 2. The position of the Zanzibar Arabs and the Northern Ireland Protestants. Each group is simultaneously dominant and precarious; each is indigenous in any reasonable sense of the term; and each depends in some substantial measure upon British support. Lofchie (1969, p. 327) writes that "[t]he Arab elite had become wholly dependent upon Great Britain for the preservation of its position in society. When Britain left, the Arabs were unable to main­ tain control on their own." A common Nationalist view, and widespread Unionist concern, is that the same would apply in the Northern Ireland case. 3. An undercurrent of intermittent, low-level political violence during rela­ tively peaceful periods. In Northern Ireland outbreaks of violence such as the sectarian clashes in Belfast in 1935, the IRA campaign of the 1950s, the Tricolour Riots of 1964, and the three murders committed by UVF members in 1966 (see, for example, Farrell, 1980) correspond to the examples from Zanzibar given above. 4. The overall character of the political struggle. In both societies there is a collision between exclusive definitions of the character of the state and the principles on which the ethnic groups should be politically incorporated within it: between Zanzibar! and African nationalism in the one, and Unionism and Catholic Nationalism in the other. Ulster Unionism being a political aspiration for a form of state and society developed in exclusive antagonism to Irish nationalism, to which no substantial fraction of Ulster Protestants adhere, this statement is valid regardless of the extent to which it is regarded as an example of "nationalism". These parallels are enough to establish the comparability of the two societies. (In terms of structure and politics Zanzibar is much more readily comparable to Northern Ireland than are the United States and South Africa, for example.) But there are also some germane differences. The most obvious is that no group within Northern Ireland corresponds to the Shirazi, over half of the population of Zanzibar. Another is that Zanzibar is largely homogeneous in terms of religion, whereas Northern Ireland is not. A third is that in Zanzibar ethnicity was to a significant extent an achieved status, whereas in Northern Ireland it is almost entirely ascribed. In the latter, one is either a Protestant or a Catholic from birth. Passing from one ethnic community to the other is rare, and there is little or no evidence that it is considered desirable or appropriate. And a fourth contrast is that Northern Ireland possesses, in both relative and absolute terms, a

13. During the crisis of 1961, the 700-strong Zanzibar police were temporarily supplemented by two General Service Companies of police and three companies of the King's African rifles, who were flown in from Kenya (Great Britain 1961, pp. 10-11,p. 18). Such reinforcements were of course unavailable during the revolution. In 1979, the Royal Ulster Constabulary, including reserves, numbered over 12,000 (Flackes, 1980, p. 200). British troop strength in Northern Ireland (including the Ulster Defence Regiment) reached a peak of 21,000 in 1972 (Barzilay, 1981, p. 237). The scale of the security operation in Northern Ireland is vastly greater than that in Zanzibar. much larger security apparatus than Zanzibar.13 The significance of these differ­ ences is assessed below.

Polarisation and Massacre In both societies the polarisation process began in a context of (at best) uneasy coexistence rather than political consensus. But in each case it did start largely with normal politics rather than violence. In Zanzibar it "resulted essentially from normal political processes acting on a structure highly conducive to racial conflict between mainland Africans and Arabs" (Kuper, 1977, p. 147), in which each group confirmed the other's threatened feeling. In Northern Ireland it also began with political aspirations unsurprising in such a plural society. The IRA campaign of 1956-1962 had collapsed for lack of support (Farrell, 1980, p. 221), and a Civil Rights Movement emerged in the late 1960s. Polarisation accelerated rapidly in each case, reaching crisis levels despite the existence of a middle ground which was substantial in numerical terms, but not well organised or obviously politically effective. The polarisation processes in the two societies are thus recognisably similar.

The same cannot be said, however, of the severity of the killing.14 At the time of writing, over 2,500 people have died in the Northern Ireland conflict since 1968. Undoubtedly, the tinge of massacre has been present on more than a few occasions, such as the La Mon and Bloody Friday bombings, Bloody Sunday, the activities of the Shankill butchers, the killing often workmen at Kingsmills in January 1976 — the list could be extended.15 Nevertheless, the Northern Ireland violence is much less severe in proportional terms. Given a population of some one and a half million, its death toll would have to reach some 25,000 in one year to match the proportion of carnage in Zanzibar. And even though the Republican

14. There has not been a great deal of comparative analysis of the violence in either Northern Ireland or Zanzibar. The main exceptions are Kuper's accounts of the process of polarisation in Algeria, Rwanda, Burundi and Zanzibar (Kuper, 1977), from which I have drawn; his discussion of Northern Ireland as a non- genocidal society in his comparative study of twentieth century genocides (Kuper, 1981); and Hewitt's analysis of the effectiveness of anti-terrorist policies in Northern Ireland and four other societies (Hewitt, 1984).

15. To give just one example, from January 1976:

The 10 victims, who all worked for Compton Spinning Mill in Glenanne, Co. Armagh, were among 12 workers travelling back to Bessbrook at around 5.30 p.m. when their minibus was stopped by a gang of between 10 and 12 armed men at the Kingsmills crossroads, just two miles from Whitecross where five Catholic men had been killed by loyalists 24 hours earlier. The gunmen, dressed in combat jackets and with their faces darkened, surrounded the bus and ordered the workmen out onto the road. They were each in turn asked their names and whether any of them were Catholics. One man replied that he was, thinking that he was the target for the gunmen. But he was told to walk the road "as fast as you can". An assortment of weapons was then used on the remaining 11 Protestant workers. The shooting lasted just a few minutes, and afterwards 10 men lay dead on the road and another was seriously injured. The sole survivor ... was shot 18 times, but miraculously lived ... (Kelly, 1986). Table 2: Agents responsible for casualties in the .Northern Ireland Conflict, 1969-1983 Agency Number Percentage

Republican Paramilitaries 1,264 54.9 Loyalist Paramilitaries 613 26.6 Security Forces 264 11.5 Not classified. 163 7.1 Total 2,304 100.1

Source: New Ireland Forum 1983, Table 4, p. 7 paramilitaries have a clear lead in responsibility for killings (see Table 2), they do not have the near monopoly that mainland Africans held in Zanzibar. The differences between the two societies noted above are not necessarily enlightening as explanations of this difference in the severity of the violence. Three of them — the large and relatively "unaffiliated" Shirazi section in Zanzibar, its common religion, and the possibility of passing — are moderating elements which do not exist in Northern Ireland. They would, if anything, lead us to expect that Zanzibar would be less at risk of massacre than Northern Ireland, yet it experienced much the higher level of violence. The fourth differ­ ence is in the size of the security forces, whose impact on the violence in Northern Ireland is controversial. We will return to it after a discussion of the impact of denial on polarisation and massacre.

IV DENIAL AND POLITICAL CONFLICT

Politics may be seen, as E.E. Schattschneider (1957, p. 935) has suggested, as the strategic activity of significantly autonomous agents whose central concern is what to do about conflict:

The grand strategy of politics deals with public policy concerning conflict. This is the policy of policies, the sovereign policy — what to do about conflict.

Schattschneider argues that this concern with conflict provides the central dynamic in the political system; and certainly in Northern Ireland and Zanzibar, the question of what to do about the ethnic and racial conflict over the very definition of the state and society is both dynamic and central. Given that conflicts of this type are notoriously difficult to resolve (e.g. Geertz, 1963; Rose, 1976), and the extreme "solution" of exterminating the opponent relatively rare, the question of what to do about it tends to recur as a chronic dilemma. In such situations, one possible orientation to the conflict is to deny it. In the most general terms, denial may be defined as the explicit rejection of a proposi­ tion which has high demonstrable truth value.16 It is not an uncommon reaction to difficult situations, and may occur on the individual level (Freud, 1946), or the collective, as in the rejection of evidence that the Holocaust was under way (Fein, 1979, Chapter 12). My focus is on its collective aspect, and there is no assumption of individual pathology in my discussion. They key question is whether there is evidence of denial in the Zanzibar and Northern Ireland conflicts, and if so, what is its impact on the processes of polarisation and massacre. Throughout the period leading up to the revolution there was in Zanzibar a widespread insistence, not justified by the historical record, that this was a virtually idyllic society, long characterised by harmonious ethnic relations. Bennett (1978, p. 251) describes the British commitment to this view:

The existing differences between the several communities of Zanzibar and Pemba, intensified by over half a century of British racially inspired policies, blocked any quick resolution of the difficulties encountered on the road to the protectorate's first election. Yet the British acted as if the incessant verbal strife between Arab and African was unnecessary. 'The Protectorate is a very happy country now', intoned Acting British Resident E. A. J. Dutton in November 1951, 'and it would be a sad pity to introduce political strife in the guise of constitutional reform.' Zanzibar's rulers apparently had become victims of their own oft-repeated dictums describing the sultanate as a country possessing model ethnic relationships.

Such views were not confined to British officials. The representatives and spokesmen of the other ethnic groups also insisted that serious violence was inconceivable. Among those giving evidence to the Commission which investi­ gated the riots of 1961 members of all the political parties were numerous and prominent. The view of Zanzibar as a country in which different races lived together in peace and harmony before 1957, apart from a few isolated incidents, "seems to have been well accepted" by them (Kuper, 1977, p. 160). More con­ cretely, the Commission was informed that "whenever the likelihood of trouble occurring on election day, 1st June, was discussed at Election Committee meet­ ings, the representatives of the political parties would invariably say, 'We Zanzibaris are peaceful people. There won't be any trouble'" (Kuper, 1977, p.

16. This formulation acknowledges that some propositions are either true or false, while others have a more relative character. 156). Denial was thus an orthodoxy shared by the conflicting parties, whatever their other disagreements. There is more evidence of denial in the Northern Ireland case, although this seems to reflect the much greater volume of research there, rather than any great difference in the prevalence of the phenomenon in the two societies. While conducting field research in West Belfast in 1972 I encountered the assertion, from both Protestant and Catholic residents of areas that had known more than their share of turbulence, that "before the Troubles" they used to coexist in harmony, paying no attention to what an individual's religious affiliation might be. Although delivered with great earnestness, these remarks posit a construc­ tion of reality radically different from that which we know to prevail in Northern Ireland, namely, a hypersensitivity to ethnic identification in face-to-face inter­ action.17 They are also at odds with the known historical record, and the ethno­ graphic evidence, formal and informal, which we possess from the "pre- Troubles" period (e.g., Harris, 1972; Harbison, 1960). A similar denial of the relevance, or even reality, of ethnic cleavage was a pro­ nounced feature of the early phase of the Civil Rights Movement. Conor Cruise O'Brien encountered it when he gave a lecture on civil disobedience at Queen's University, Belfast, in late October, 1968. The audience reaction was negative:

What I was criticized for, quite heatedly, was for mentioning an aspect of reality: the existence of two separate communities, Catholics and Pro­ testants. This was held to be "irrelevant", a favourite all-purpose student knock­ out word at the time, in Belfast as well as in New York. "Religion", one student said, "is a red herring", I said if so it was a red herring about the size of a whale. "No, no", said another student, "no one in Ulster is the least bit interested in religion", "Not even in Sandy Row?" I asked. Another student said he himself came from Sandy Row, and could report "that no one there cared whether a man was a Catholic or Protestant". This thumping lie was loudly applauded (O'Brien, 1974, p. 149-150).18 While these examples on their own are not enough to establish that denial is widespread, other research confirms its apparently systematic character. Nelson (1975, p. 160) reports the results of "about fifty" informal interviews with Protestant respondents:

17. Knowledge of ethnic identity is treated as essential in day-to-day negotiation of the social world in Northern Ireland. See for example Burton's discussion (1978, p. 4, p. 37 and passim) of "telling".

18. Only those unfamiliar with Northern Ireland politics will require the information that Sandy Row, a Protestant working-class district near to the University, is renowned as a bastion of ethnic solidarity. I opened my discussion of "discrimination" with respondents with the question: there's been a lot of talk in Northern Ireland about "discrimina­ tion" against Catholics. Do you think this existed or not? ... the reaction of the majority was denial But the denial was almost without exception qualified during the discussion which followed, so that virtually all respon­ dents had admitted to some forms of anti-Catholic discrimination at least by the conclusion of the interview.

In a study of intimidation, Darby (1985) reports 4 denying mythologies as particularly common in his study area: a nostalgia myth, which looks back to the early 1960s as a golden age of harmony "when people lived together in peace and happiness without a thought about each others' religion";19 an invasion myth, according to which all personal acts of intimidation were carried out by outsiders to the area; a conspiracy myth that intimidation was highly organised, usually by a paramilitary organisation; and a vandalism myth that it was only members of "the other side" who vandalised homes which they were forced by intimidation to leave. Darby also gives a quintessential example of a young woman who refused to accept that any Protestants had been forced to leave the area in which she lived. She was living at the time in a house from which Pro­ testant residents had been forced to flee. There is also evidence of denial that members of one's own ethnic group were responsible for killings. Nelson (1984, p. 122), for example, in discussing the first book to publicise the fact that in 1972-1973 many sectarian assassinations were carried out by Protestant murder squads (Dillon and Lehane, 1973), makes the following comments:

People would speak with distaste or condemnation about sadistic killings, sometimes openly, in conversations on the street, or privately in areas like Glencairn where hooded bodies were found with some frequency.20 But even in areas where militants came from, many people simply would not believe Protestants had done the killings at all. Outside these areas, dis­ belief was still more common. Of course, parties to wars everywhere are reluctant to believe their own side's atrocities: what was striking is the length of time the belief persisted. It is interesting that while few para- militants denied the substance (if not the detail) of Dillon and Lehane's book was accurate, many Protestant civilians and politicians seemed unable to accept its conclusions several years later.

19. This replicates my finding noted on page 304.

20. Note that there is an implication here that people did not speak with distaste or condemnation about "non sadistic" killings. My ongoing research shows that most killings were "non-sadistic". Since Nelson's and Darby's subjects are mainly working class, it is clear that denial is not an elite prerogative. Nor are outsiders immune to it. In the 26 counties of the Irish Republic the denying principle that there can be no insuperable cleavage between the ethnic groups in Ireland is a central Nationalist tenet. The book in which O'Brien's vignette appeared is an onslaught against denial in Ireland as a whole, and as such a threat to the legitimating mythology of the Irish state (Brown, 1981, pp. 283-291). And many British politicians, particularly members of the Labour Party, have approached Northern Ireland with the perspective, articulated in policy documents and else­ where (e.g., British Labour Party, 1981), that its politics are primarily class politics. This view, an ideological and cultural product of British society, is conducive to denial when applied in a political milieu, such as Northern Ireland, in which class conflict is not dominant (Thompson, 1979, 1980).

V THE UNASSAILABILITY OF DENYING ROUTINES

While there is evidence that denial is a widespread orientation, its significance remains to be established. To the extent that it is easily challenged and over­ ridden, it is not likely to be consequential. The continuation of O'Brien's account of his lecture at Queen's University in 1968 is suggestive here. He describes the reaction to the stout declaration that the residents of Sandy Row were indifferent to ethnic identity:

Hearing the applause, the speaker seemed to have some doubts about what he had uttered. He added, hesitatingly, "They just come out when the drums beat, you know." We knew. But why did they come out when the drums beat? Someone said, rather dreamily: "Well, anyone would." The more orthodox answers soon followed; people came because they had been brain-washed, duped by the bosses. But were the bosses really so clever, the workers so dumb? And why should Protestant workers be dumber than Catholic workers? Well, the Protestant workers did have some privileges, small ones, which the bosses encouraged them to over-value. In any case, it was waste [sic] of time to speculate why these things should be so. The important thing was action; solidarity of Protestant and Catholic workers, dictated by basic, common class interest, would grow out of the struggle itself, dissipating false consciousness, while destroying the structures which perpetuated it. It was easy to know that would not happen ... (O'Brien quotes survey evidence [Rose 1971] that ethnic loyalties in Northern Ireland were far stronger than those reflecting class.) The speaker from Sandy Row, having second thoughts, "admits" what every­ one present is routinely aware of— that the residents of Sandy Row are highly responsive to ethnic appeals ("'They just come out when the drums beat, you know.' We knew"). The members of the audience then begin to conduct the discussion on the basis of the premise that they had just attacked O'Brien for articulating, i.e., that their society is in fact riven by ethnic cleavage. This indicates that their starting position was indeed denial. The fact that they do not, as a group, critically attack the speaker for his volte-face, but co-operate in it, without apparent strain, suggests that his reversal is a manoeuvre with which they are quite familiar. The next step is another defensive and denying enterprise, this time in the form of offering explanations of the ethnic cleavage in terms of class. The theories offered in explanation are inadequate as an attempt at intellectual under­ standing, as O'Brien's critique, and other evidence,21 make clear. But his objec­ tions are dismissed, and the need for the course of action based on the initial denial is affirmed. Such switching between contradictory positions, without challenge by collaborators who share an inability or unwillingness to acknowledge the problems of ethnic accommodation, illustrates the co­ operative and supra-individual element in denial. The intervention by O'Brien, an outsider both sufficiently knowledgeable and sufficiently belligerent to issue a public challenge, was exceptional — a fact which itself implies that denial was the modal orientation of his audience. Those who intervene in such a deflating way can hardly expect affectionate embrace, and his audience exorcised him as quickly as possible. In so doing, they collec­ tively regressed to the original premise which they had temporarily abandoned under his first attack. The overall response to challenge was a reaffirmation of denial. The exchange as a whole suggests the unassailability of entrenched deny­ ing routines. Since they are not readily amenable to successful challenge, they may be regarded as part of a standard repertoire of techniques of collective action (Tilly, 1978, pp. 151-156). As such, they may have significant conse­ quences, both intended and unintended.

VI THE CONSEQUENCES OF DENIAL

In 1961 the representatives of the political parties in Zanzibar held that, given the peaceable disposition of their constituents, violence at the polls was incon­ ceivable. As a result, they did not co-operate in creating a structure for conflict management. British officials were in a poor position to persuade them to do so,

21. Detailed re-analyses of Rose's data which confirm this provided by Thompson, (1983) and Kelley and McAllister, (1984). See also Whyte, (1978). since they officially subscribed to the same view. The outcome was that 68 people died in the ensuing violence. In Northern Ireland denial, whether the vulgar Marxisant variant of the radical elements of the Civil Rights Movement, or the other types discussed above, does not change the political reality that ethnic cleavage predominates, but leads to avoidance of that reality, and contributes to the process of ethnic polarisation. The same holds for denial by involved outsiders. A result of the South's tenet that there can in principle be no insuperable cleavage between ethnic groups in Ireland is that any limited co-operation is interpreted as both a practical and symbolic demonstration of movement towards an all-Ireland unit, rather than an end in itself. This helps to render such limited accommodation difficult or impossible. British and other orientations to the Northern Ireland problem which treat it in class terms have the same consequences, as was demon­ strated for example in the Ulster Worker's Council Strike of 1974 (Thompson, 1980). There is, then, support for the proposition that to the extent that mobilisation in movements which seek, or claim to seek, political accommodation between ethnic sections is informed and structured by denial, the prognosis is likely to be failure, possibly leading to further polarisation. The failure does not arise because accommodation which dampens polarisation and leads to de-escalation is inherently impossible even in societies like Zanzibar and Northern Ireland. It occurs because such accommodation can hardly take place if the participants are incapable of articulating in public a recognition of the real nature of the society, the cleavages within it, and the difficult political problems which they generate. It is not so much what people "know" privately that is significant: it is what it is possible for them to accept publicly. The former is undoubtedly important, but the latter is of central concern because solutions, compromises and accommodations to overt ethnic conflicts must be publicly argued, defended and accepted.

VII THREE EXPANSIONS

Within-group denial In Northern Ireland,22 contact between members of different ethnic groups is rare in childhood, most political socialisation and education occurs within the ethnic group, and adults are uncomfortable in discussion of central political cleavages with out-group members even in relatively peaceful periods (Murray 1983; Darby et al. 1977; Devlin 1969; McCann 1980; Harbinson 1960; Harris

22. No substantial literature on Zanzibar deals with the topics mentioned in this paragraph. 1972). We may therefore suggest that it is largely within the group that the repertoire of entrenched denying routines is created. By the nature of denying statements, and the aggressiveness with which they are delivered,23 they are difficult to challenge. One tends to let them pass, there­ by reinforcing them. This alerts us to one of their key uses in internal "discus­ sions": as a way of suppressing heterodoxy. If unrealistic and denying definitions of the situation are routinely enforced within the group, and withdrawal from them is defined as a desertion of group solidarity, as disloyalty, then realistic within-group discussion of political compromise and accommodation can be expected to be both difficult and rare. In the short term, this has the result that we have already noted: no discussion of the basis of conflict, or of what might be an accommodation to some of the concerns of the other community, can begin. The long-term outcome is the refurbishment, or institutionalisation, of a. denying culture, i.e., the construction, codification, and refinement of a standard repertoire of entrenched denying routines. A further long-term result may be that when the conflict eventually becomes so overt and severe that members of the society wish to regulate it — for example when polarisation reaches an extreme point, or killing threatens to get "out of hand" — they are incapable of doing so. Thus we may partly explain the apparent inability of the major partici­ pants in the Northern Ireland conflict to actively engage in productive political negotiation.

The consensus of opponents: between-group denial Despite the above — and partly because of it — diverse elements who are otherwise fierce political opponents may converge or collaborate in denial. Thus in Zanzibar the British and the emergent nationalists concurred in the public definition of the society as one of idyllic harmony in its racial relations; in 1972 both my Protestant and Catholic respondents insisted on the unproblematic bliss of the pre-Troubles period. Such consensus in denial by antagonists has special significance for polarisation precisely because the parties to it are opponents. This very fact gives their denying consensus a curious, malign legitimacy. For if members of two groups known for their antagonism share a relatively benign analysis of the conflict, how can outsiders challenge it? In Zanzibar, as the conflict became more overt, the British took strenuous security measures.24 Kuper (1977, p. 156-157) argues convincingly that without these, violence would have been more severe. These actions show both that violence was for the British eminently conceivable, and that in a crisis their pre­ vious denying orientation did not prevent them from autonomous preventive

2.3. I was struck by this during my ficldwork in 1972.

24. See note 13. action; the independence of ideology and action is (once more) demonstrated. Nevertheless, they were unable to persuade the other parties to join them in co­ operative restraining measures, which could reasonably be expected to further reduce the level of violence. This may have been partly because the British had promoted the denying position for some time, and had not previously facilitated political representation for and co-operation between the subordinate groups. They were therefore, as I suggested above, badly placed to argue for new action which recognised their antagonism. This suggests a sense in which denial among opponents is more conducive to polarisation than denial among allies. Let us assume that, up to a certain level, polarisation can be contained by autonomous action, but beyond that, co­ operation between opponents is required. This seems quite realistic. Since auto­ nomous preventive action is relatively unproblematic, it follows that the polarisation process may be relatively slow up to that point, even if there is denial. But beyond it, an acceleration in the rate of polarisation may be expected, since the necessary co-operation is less likely to be forthcoming.

The genesis and constraint of massacre One analysis of the genesis of massacre in Northern Ireland on which our comparison casts some light is O'Brien's recently reasserted "malign model" (O'Brien 1985). According to it, one reason for recurring violence in Northern Ireland, and one which suggests that still more serious violence is possible in the future, is the denial by both constitutional Nationalism and physical force Republicanism of the depth and nature of the gulf between them and the Protestant community. Regarding Nationalists, O'Brien writes of: —

the traditional — and apparently incurable — pan-Nationalist incapacity to see the Northern Unionists as an autonomous force, with deep convic­ tions of its own. For more than a century now it has been Nationalist doctrine that, once Dublin and London are in accord, the Ulster Unionists will have no choice but to toe the line. That this doctrine does not fit the facts was demon­ strated in 1912-1914 and again in 1974. But the doctrine, although objec­ tively refuted, remains subjectively intact.

The Republican variant, that the British are the cause of political violence in Ireland, that their departure is a precondition for a solution, and that in their absence it will not get out of control, is a sweeping denial of the severity of the internal ethnic cleavage. In both cases, the efforts to impose all-Ireland institu­ tions can be expected to lead to an increase in violence, and the nearer they come to "success", the greater the likely increase in violence. Since the malign scenario was enacted in Zanzibar, and since our compara­ tive analysis finds considerable similarities between it and Northern Ireland, the reasonable conclusion is that the malign model is relevant to the latter case. To reject it, plausible restraining factors, present there and absent in Zanzibar, must be identified. The obvious possibility suggested by the analysis is the continuing presence of the security forces. One relevant element seems simply to be the size of the force deployed. Thus Kuper concludes that in Zanzibar, had the British not introduced more troops in 1961, the death toll in the riots of that year would have been higher. One might argue that this is simply a misinterpretation, but this seems untenable, since soon after the withdrawal of the British, and their security resources, there was massacre. The effects of a larger security apparatus, however, are conditional on its orientation to outbreaks of killing. This can vary independently of size, as the Northern Ireland case shows. In the 1972-1973 period, police spokesmen offi­ cially treated the sectarian assassinations, most of which were committed by Protestants, as "motiveless murders." This orientation changed over time, how­ ever, and by the late 1970s and early 1980s the police were pursuing Protestant paramilitary murderers with much greater vigour. This is evident in police statistics and reports (e.g., Royal Ulster Constabulary 1983), the use of super­ grasses (or converted terrorists, in official terminology) against Loyalist as well as Republican paramilitaries, and the greatly increased antagonism of Loyalist groups to the police. The reasons for this shift, which may include political direc­ tion from the top, and increasing police professionalisation and organisational autonomy, merit investigation, but the change in orientation, and the decline in the assassinations by Protestant paramilitaries, is not in doubt. While other factors are involved, it is hard to doubt that the change in the attitude of the police was one reason for this downturn. From a policy perspective, the orientation of the security forces to such killing is independent of the imposition of all-Ireland institutions. Whether the current police stance, and its restraining effects, will continue, and whether these will be outweighed by the polarising effects of efforts to impose all-Ireland institutions, is a matter of no small importance.

VIII CONCLUSION

On the general level, this paper has advanced two propositions. The first is that denial is one possible response to problematic political, as toother, conflict. The second is that if actors cannot publicly acknowledge a conflict and its character, they cannot negotiate or regulate it efficiently. This approach can be further developed in many ways. Explorations of the conditions under which denial is most likely to occur, the particular aspects of a conflict which are most likely to be denied, and who is most likely to engage in denial, are possible. Since denial is a response to "trouble", two possible general hypotheses are that it is the most problematic aspects of a conflict which are most likely to be denied, and that it is those who are least equipped with the ability to negotiate them who are most likely to deny them. These can be treated, in keeping with the level of analysis here, as propositions about collective and cultural resources. In addi­ tion, denial is only one possible orientation to conflict, and not even the only one which leads to a failure to address it. A comprehensive typology specifying all the possible orientations, and their likely consequences, might therefore be sought. Theories accounting for the different orientations, and changes in them, could then be formulated. In terms of specific applications in this paper, enough has perhaps been said to suggest the utility of the two general propositions in explaining rapid polarisa­ tion and the genesis of massacre in the cases studied. One obvious problem which deserves a final comment is that of how to establish empirically that denial is occurring. In principle, it is possible to decide whether or not a disputed proposi­ tion about social reality has demonstrable validity: if it does, we have denial. In practice, however, given the nature of denial, the prospect for polemics is high. But help may be at hand from a somewhat surprising quarter. The examples considered above suggest that actors themselves often display some level of awareness that they are in fact engaging in denial. Thus, as Kuper (1977, p. 163-164) notes, the Riot Commissioners in Zanzibar were not surprised that the racial and ethnic harmony to which those appearing before them testified deteriorated rapidly, which suggests an awareness that discord lay only a short way below the surface of the society. Nelson needed only to let her Protestant respondents continue to talk, and their admissions that discrimination existed consistently emerged (although one senses that if asked the same question again later, they would be likely to begin with a similar denying response). We will be on solid ground if we adopt the following criteria for establishing denial: either firm, publicly available evidence that the disputed proposition is invalid; or retractions such as those elicited by Nelson.

REFERENCES

BARZILAY, DAVID, 1981. The British Army in Ulster, volume 4. Belfast: Century. BENNETT, NORMAN R., 1978. A History of the Arab State of Zanzibar, London: Methuen. BRITISH LABOUR PARTY, 1981. Northern Ireland: Statement by the National Executive Committee, London: Labour Party. BROWN, TERENCE, 1981. Ireland: A Social and Cultural History 1922-1979, Glasgow: Fontana. BUCKLAND, PATRICK, 1981. A History of Northern Ireland, Dublin: Gill and Macmillan. BURTON, FRANK, 1978. The Politics of Legitimacy: Struggles in a Belfast Community, London: Routledge and Kegan Paul. DARBY, JOHN, 1976. Conflict in Northern Ireland: The Development of a Divided Community, New York: Barnes and Noble. DARBY, JOHN, el al. 1977. Education and Community in Northern Ireland: Schools Apart? Coleraine: New University of Ulster. DARBY, JOHN, (ed.), 1983. Northern Ireland: The Background lo the Conflict, Belfast: Appletree and Syracuse: Syracuse University Press. DARBY, JOHN, 1985. Intimidation and the Control of Inter-group conflict in Northern Ireland, Ph.D. Dissertation, Faculty of Health and Social Sciences, University of Ulster. DESJARDINS, THIERRY, 1976. he Martyre du Liban, Paris: Plon. DEVLIN, BERNADETTE, 1969. The Price of My Soul, London: Pan. DILLON, MARTIN and DENIS LEHANE, 1973. Political Murder in Northern Ireland, Harmonds- worth: Penguin. FARRELL, MICHAEL, 1980. Northern Ireland: The Orange Slate, Second, revised edition. London: Pluto. FLACKES, W. D., 1980. Northern Ireland: A Political Directory, 1968-79, Dublin: Gill and Mac- millan and New York: St Martin's. FREDRICKSON, GEORGE M., 1981. White Supremacy: A Comparative. Study in American and South African History, New York: Oxford. FREUD, ANNA, 1946. The Ego and the Mechanisms of Defence, New York: International Publishers. GEERTZ, CLIFFORD, 1963. "The integrative revolution: Primordial sentiments and civil polit ics in the new states", pp. 105-157 in Old Societies and New Slates: The Qiiest for Modernity in Asia and Africa, New York: Free Press. GREAT BRITAIN, 1961. Report of a Commission of Inquiry into Disturbances in Zanzibar during June 1961, Colonial No. 353. London: Her Majesty's Stationery Office. HARBINSON, R., 1960. No Surrender: An Ulster Childhood, London: Faber. HARRIS, ROSEMARY, 1972. Prejudice and Tolerance in Ulster: A study of neighbours and 'strangers' in a border community, Manchester: Manchester University Press, and Totawa, N. ].: Rowmanand Littlefield. HEWITT, CHRISTOPHER, 1984. The Effectiveness of Anti-Terrorist Policies, New York: Univer­ sity Press of America. HILBERG, RAUL, 1961. The Destruction of the European Jews, Chicago: Quadrangle Books. KELLEY, JONATHAN and IAN McALLlSTER, 1984. "The Genesis of Conflict: Religion and Status Attainment in Ulster, 1968.'' Sociology, 18: 171-190. KELLY, MARY, 1986. "Service to mark Kingsmills Massacre", Irish Times, January 4, p. 17. KHARUSI, A. S., 1967. Zanzibar - Africa's First Cuba, Richmond: Foreign Affairs Publishing. KUPER, LEO, 1977. The Pity of It All: Polarization of Ethnic and Racial Relations, London: Duck­ worth. KUPER, LEO, 1981. Genocide: Its Political Use in the Twentieth Century, Great Britain: Penguin Books. KUPER, LEO and M. G. SMITH (eds.), 1969. Pluralism in Africa. Berkeley: University of Cali­ fornia Press. LIJPHART, AREND, 1971. "Comparative Politics and the Comparative Method", American Political Science Review 65:682-693. LOFCHIE, MICHAEL F., 1965. Zanzibar: Background to Revolution. Princeton: Princeton Univer­ sity Press. LOFCHIE, MICHAEL F., 1969. "The Plural Society in Zanzibar", in Leo Kuper and M. G. Smith (eds.) Pluralism in Africa, Berkeley: University of California Press, pp. 283-328. MARTIN, ESMOND BRADLEY, 1978. Zanzibar: Tradition and Revolution, London: Hamish Hamilton. McCANN, EAMONN, 1980. War and an Irish Town, Second Edition. London: Pluto. MURRAY, DOMINIC, 1983. "Schools and Conflict", in John Darby (ed.) Northern Ireland: The Background to the Conflict, Belfast: Appletree and Syracuse: Syracuse University Press, pp. 136- 150. NELSON, SARAH, 1975. "Protestant 'Ideology' Considered: The Case of'Discrimination'", in Ivor Crewe (ed.), British Political Sociology Yearbook, Volume 2: The Politics of Race, pp. 155-187. NELSON, SARAH, 1984. Ulster's Uncertain Defenders: Protestant Political, Paramilitary and Community Groups in the Northern Ireland Conflict, Belfast: Appletree. NEW IRELAND FORUM, 1983. The Cost of Violence arising from the Northern Ireland Crisis since 1969, Dublin: The Stationery Office. O'BRIEN, CONOR CRUISE,' 1974. States of Ireland, St. Alban's: Panther. O'BRIEN, CONOR CRUISE, 1985. "Nationalist Doctrine does not fit the facts", Irish Times, September 10. OKELLO, JOHN, 1967. Revolution in Zanzibar, Nairobi: East African Publishing House. ROSE, RICHARD, 1971. Governing Without Consensus: An Irish Perspective, Boston: Beacon. ROSE, RICHARD, 1976. Northern Ireland: A Time of Choice, Washington: American Enterprise Institute for Public Policy Research. ROYAL ULSTER CONSTABULARY, 1983. Chief Constable's Report 1982, Belfast: The Police Authority for Northern Ireland. SCHATTSCHNEIDER, E. E., 1957. "Intensity, Visibility, Direction and Scope", American Political. Science Review, 51: 933-942. STEWART, A. T. Q., 1977. The Narrow Ground: Aspects of Ulster 1609-1969, London: Faber. THOMPSON, JOHN L. P., 1979. Dual Incorporation in Northern Ireland: A Theory of Social and Political Structure, Ph.D. dissertation, University of California, Los Angeles. THOMPSON, JOHN L. P., 1980. "Structural Ambiguity in Northern Ireland: the Ulster Workers' Council Strike of 1974", Paper read at the Meetings of the American Council for Irish Studies, Washington D.C. THOMPSON, JOHN L. P., 1983. "The Plural Society Approach to Class and Ethnic Political Mobilization", Ethnic and Racial Studies, 6 No. 2 (April): 127-153. THOMPSON, JOHN L. P., 1985a. "Analyzing Political Killing: Concepts, Measures, and the Problem of Intent", Paper presented at the Annual Conference of the British Sociological Assoc­ iation, Hull, England, April. THOMPSON, JOHN L. P., 1985b. "Crime, Social Control, and Trends in Political Killing", Paper presented at the Annual Meetings of the Sociological Association of Ireland, Belfast. TILLY, CHARLES, 1978. From Mobilization To Revolution, Reading, MA.: Addison-Wesley. VAN DEN BERGHE, PIERRE, 1978. Race and Racism: A Comparative Perspective, New York: Wiley. WHYTE, JOHN, 1978. "Interpretations of the Northern Ireland Problem: An Appraisal", The Economic and Social Review, 9: 257-282. ZANZIBAR PROTECTORATE, 1953. Notes on the Census of the Zanzibar Protectorate 1948, Zan­ zibar: Government Printer.